melashri/SUSY
Watch
1
0
Fork
You've already forked SUSY
 0
mirror of https://github.com/MohamedElashri/SUSY.git synced 2026-11-05 03:45:24 -05:00
Code Issues Projects Releases Packages Wiki Activity
Experimenting Machine Learning Techniques on SUSY dataset.
17 commits 1 branch 1 tag 338 KiB
  • Jupyter Notebook 100%
Find a file
Exact
Repository files (latest commit first)
Filename Latest commit message Latest commit date
Mohamed Elashri 80e07265bb Created using Colaboratory
2021-03-02 22:05:55 -05:00
.whitesource Add .whitesource configuration file 2021-03-01 21:25:31 +00:00
docker-lock.json Updated Lockfile 2021-03-01 21:51:58 +00:00
LICENSE Create LICENSE 2021-03-02 12:04:15 -05:00
README.md Update README.md 2021-03-02 14:48:39 -05:00
SUSY.ipynb Created using Colaboratory 2021-03-02 11:37:39 -05:00
SUSY_DNN.ipynb Created using Colaboratory 2021-03-02 22:05:55 -05:00
SUSY_NN.ipynb Created using Colaboratory 2021-03-02 14:40:27 -05:00

README.md

SUSY

Experimenting Machine Learning Techniques on SUSY dataset.

Dataset

I'm using a Monte Carlo generated SUSY dataset. This dataset introduced in [1], The supersymmetry data set consists of 5,000,000 Monte-Carlo samples of supersymmetric and non-supersymmetric collisions with 18 features. The signal process is the production of electrically-charged supersymmetric particles which decay to W bosons and an electrically-neutral supersymmetric particle that is invisible to the detector.

The first 8 features are considered low which means that they can directly be measured from collisions. The other 10 are high-level featured which means that they are reconstructed (reconstruction in particle physics means hand constructed based on our physics knowledge). These high-level features are known to be important in our classification problem.

SUSY Dataset can be downloaded from this Github repository releases.

Dataset Source: UCI

Project Purpose

I want to have a real exercise about applying ML Algorithms on real particle physics dataset. Super-Symmetry is one of important searches conducted now and the paper presenting the original anslysis keeps room for improvment. My goal is to play with the dataset and try to get efficent and more accurate classification. Also the computation time and resources are always constrained so I will try to optimize that for our additional gain, if any.

References

[1] Baldi, P., Sadowski, P. & Whiteson, D. Searching for exotic particles in high-energy physics with deep learning. Nat Commun 5, 4308 (2014). https://doi.org/10.1038/ncomms5308